Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Content-Type: text/plain;
        charset=us-ascii
Mime-Version: 1.0 (Mac OS X Mail 13.0 \(3594.4.18\))
Subject: Re: [PATCH 14/15] mm: Align THP mappings for non-DAX
From:   William Kucharski <william.kucharski@oracle.com>
In-Reply-To: <20191001104558.rdcqhjdz7frfuhca@box>
Date:   Tue, 1 Oct 2019 05:21:26 -0600
Cc:     Matthew Wilcox <willy@infradead.org>,
        linux-fsdevel@vger.kernel.org, linux-mm@kvack.org,
        linux-kernel@vger.kernel.org
Content-Transfer-Encoding: quoted-printable
Message-Id: <A935F599-BB18-40C3-90DD-47B7700743D6@oracle.com>
References: <20190925005214.27240-1-willy@infradead.org>
 <20190925005214.27240-15-willy@infradead.org>
 <20191001104558.rdcqhjdz7frfuhca@box>
To:     "Kirill A. Shutemov" <kirill@shutemov.name>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


> On Oct 1, 2019, at 4:45 AM, Kirill A. Shutemov <kirill@shutemov.name> =
wrote:
>=20
> On Tue, Sep 24, 2019 at 05:52:13PM -0700, Matthew Wilcox wrote:
>>=20
>> diff --git a/mm/huge_memory.c b/mm/huge_memory.c
>> index cbe7d0619439..670a1780bd2f 100644
>> --- a/mm/huge_memory.c
>> +++ b/mm/huge_memory.c
>> @@ -563,8 +563,6 @@ unsigned long thp_get_unmapped_area(struct file =
*filp, unsigned long addr,
>>=20
>> 	if (addr)
>> 		goto out;
>> -	if (!IS_DAX(filp->f_mapping->host) || =
!IS_ENABLED(CONFIG_FS_DAX_PMD))
>> -		goto out;
>>=20
>> 	addr =3D __thp_get_unmapped_area(filp, len, off, flags, =
PMD_SIZE);
>> 	if (addr)
>=20
> I think you reducing ASLR without any real indication that THP is =
relevant
> for the VMA. We need to know if any huge page allocation will be
> *attempted* for the VMA or the file.

Without a properly aligned address the code will never even attempt =
allocating
a THP.

I don't think rounding an address to one that would be properly aligned =
to map
to a THP if possible is all that detrimental to ASLR and without the =
ability to
pick an aligned address it's rather unlikely anyone would ever map =
anything to
a THP unless they explicitly designate an address with MAP_FIXED.

If you do object to the slight reduction of the ASLR address space, what
alternative would you prefer to see?

    -- Bill=